The BPF program loading process performed by libbpf is quite complex
and consists of the following steps:
"open" phase:
- parse elf file and remember relocations, sections
- collect externs and ksyms including their btf_ids in prog's BTF
- patch BTF datasec (since llvm couldn't do it)
- init maps (old style map_def, BTF based, global data map, kconfig map)
- collect relocations against progs and maps
"load" phase:
- probe kernel features
- load vmlinux BTF
- resolve externs (kconfig and ksym)
- load program BTF
- init struct_ops
- create maps
- apply CO-RE relocations
- patch ld_imm64 insns with src_reg=PSEUDO_MAP, PSEUDO_MAP_VALUE, PSEUDO_BTF_ID
- reposition subprograms and adjust call insns
- sanitize and load progs
During this process libbpf does sys_bpf() calls to load BTF, create maps,
populate maps and finally load programs.
Instead of actually doing the syscalls generate a trace of what libbpf
would have done and represent it as the "loader program".
The "loader program" consists of single map with:
- union bpf_attr(s)
- BTF bytes
- map value bytes
- insns bytes
and single bpf program that passes bpf_attr(s) and data into bpf_sys_bpf() helper.
Executing such "loader program" via bpf_prog_test_run() command will
replay the sequence of syscalls that libbpf would have done which will result
the same maps created and programs loaded as specified in the elf file.
The "loader program" removes libelf and majority of libbpf dependency from
program loading process.
kconfig, typeless ksym, struct_ops and CO-RE are not supported yet.
The order of relocate_data and relocate_calls had to change, so that
bpf_gen__prog_load() can see all relocations for a given program with
correct insn_idx-es.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210514003623.28033-15-alexei.starovoitov@gmail.com
In order to be able to generate loader program in the later
patches change the order of data and text relocations.
Also improve the test to include data relos.
If the kernel supports "FD array" the map_fd relocations can be processed
before text relos since generated loader program won't need to manually
patch ld_imm64 insns with map_fd.
But ksym and kfunc relocations can only be processed after all calls
are relocated, since loader program will consist of a sequence
of calls to bpf_btf_find_by_name_kind() followed by patching of btf_id
and btf_obj_fd into corresponding ld_imm64 insns. The locations of those
ld_imm64 insns are specified in relocations.
Hence process all data relocations (maps, ksym, kfunc) together after call relos.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210514003623.28033-12-alexei.starovoitov@gmail.com
This adds functions that wrap the netlink API used for adding, manipulating,
and removing traffic control filters.
The API summary:
A bpf_tc_hook represents a location where a TC-BPF filter can be attached.
This means that creating a hook leads to creation of the backing qdisc,
while destruction either removes all filters attached to a hook, or destroys
qdisc if requested explicitly (as discussed below).
The TC-BPF API functions operate on this bpf_tc_hook to attach, replace,
query, and detach tc filters. All functions return 0 on success, and a
negative error code on failure.
bpf_tc_hook_create - Create a hook
Parameters:
@hook - Cannot be NULL, ifindex > 0, attach_point must be set to
proper enum constant. Note that parent must be unset when
attach_point is one of BPF_TC_INGRESS or BPF_TC_EGRESS. Note
that as an exception BPF_TC_INGRESS|BPF_TC_EGRESS is also a
valid value for attach_point.
Returns -EOPNOTSUPP when hook has attach_point as BPF_TC_CUSTOM.
bpf_tc_hook_destroy - Destroy a hook
Parameters:
@hook - Cannot be NULL. The behaviour depends on value of
attach_point. If BPF_TC_INGRESS, all filters attached to
the ingress hook will be detached. If BPF_TC_EGRESS, all
filters attached to the egress hook will be detached. If
BPF_TC_INGRESS|BPF_TC_EGRESS, the clsact qdisc will be
deleted, also detaching all filters. As before, parent must
be unset for these attach_points, and set for BPF_TC_CUSTOM.
It is advised that if the qdisc is operated on by many programs,
then the program at least check that there are no other existing
filters before deleting the clsact qdisc. An example is shown
below:
DECLARE_LIBBPF_OPTS(bpf_tc_hook, .ifindex = if_nametoindex("lo"),
.attach_point = BPF_TC_INGRESS);
/* set opts as NULL, as we're not really interested in
* getting any info for a particular filter, but just
* detecting its presence.
*/
r = bpf_tc_query(&hook, NULL);
if (r == -ENOENT) {
/* no filters */
hook.attach_point = BPF_TC_INGRESS|BPF_TC_EGREESS;
return bpf_tc_hook_destroy(&hook);
} else {
/* failed or r == 0, the latter means filters do exist */
return r;
}
Note that there is a small race between checking for no
filters and deleting the qdisc. This is currently unavoidable.
Returns -EOPNOTSUPP when hook has attach_point as BPF_TC_CUSTOM.
bpf_tc_attach - Attach a filter to a hook
Parameters:
@hook - Cannot be NULL. Represents the hook the filter will be
attached to. Requirements for ifindex and attach_point are
same as described in bpf_tc_hook_create, but BPF_TC_CUSTOM
is also supported. In that case, parent must be set to the
handle where the filter will be attached (using BPF_TC_PARENT).
E.g. to set parent to 1:16 like in tc command line, the
equivalent would be BPF_TC_PARENT(1, 16).
@opts - Cannot be NULL. The following opts are optional:
* handle - The handle of the filter
* priority - The priority of the filter
Must be >= 0 and <= UINT16_MAX
Note that when left unset, they will be auto-allocated by
the kernel. The following opts must be set:
* prog_fd - The fd of the loaded SCHED_CLS prog
The following opts must be unset:
* prog_id - The ID of the BPF prog
The following opts are optional:
* flags - Currently only BPF_TC_F_REPLACE is allowed. It
allows replacing an existing filter instead of
failing with -EEXIST.
The following opts will be filled by bpf_tc_attach on a
successful attach operation if they are unset:
* handle - The handle of the attached filter
* priority - The priority of the attached filter
* prog_id - The ID of the attached SCHED_CLS prog
This way, the user can know what the auto allocated values
for optional opts like handle and priority are for the newly
attached filter, if they were unset.
Note that some other attributes are set to fixed default
values listed below (this holds for all bpf_tc_* APIs):
protocol as ETH_P_ALL, direct action mode, chain index of 0,
and class ID of 0 (this can be set by writing to the
skb->tc_classid field from the BPF program).
bpf_tc_detach
Parameters:
@hook - Cannot be NULL. Represents the hook the filter will be
detached from. Requirements are same as described above
in bpf_tc_attach.
@opts - Cannot be NULL. The following opts must be set:
* handle, priority
The following opts must be unset:
* prog_fd, prog_id, flags
bpf_tc_query
Parameters:
@hook - Cannot be NULL. Represents the hook where the filter lookup will
be performed. Requirements are same as described above in
bpf_tc_attach().
@opts - Cannot be NULL. The following opts must be set:
* handle, priority
The following opts must be unset:
* prog_fd, prog_id, flags
The following fields will be filled by bpf_tc_query upon a
successful lookup:
* prog_id
Some usage examples (using BPF skeleton infrastructure):
BPF program (test_tc_bpf.c):
#include <linux/bpf.h>
#include <bpf/bpf_helpers.h>
SEC("classifier")
int cls(struct __sk_buff *skb)
{
return 0;
}
Userspace loader:
struct test_tc_bpf *skel = NULL;
int fd, r;
skel = test_tc_bpf__open_and_load();
if (!skel)
return -ENOMEM;
fd = bpf_program__fd(skel->progs.cls);
DECLARE_LIBBPF_OPTS(bpf_tc_hook, hook, .ifindex =
if_nametoindex("lo"), .attach_point =
BPF_TC_INGRESS);
/* Create clsact qdisc */
r = bpf_tc_hook_create(&hook);
if (r < 0)
goto end;
DECLARE_LIBBPF_OPTS(bpf_tc_opts, opts, .prog_fd = fd);
r = bpf_tc_attach(&hook, &opts);
if (r < 0)
goto end;
/* Print the auto allocated handle and priority */
printf("Handle=%u", opts.handle);
printf("Priority=%u", opts.priority);
opts.prog_fd = opts.prog_id = 0;
bpf_tc_detach(&hook, &opts);
end:
test_tc_bpf__destroy(skel);
This is equivalent to doing the following using tc command line:
# tc qdisc add dev lo clsact
# tc filter add dev lo ingress bpf obj foo.o sec classifier da
# tc filter del dev lo ingress handle <h> prio <p> bpf
... where the handle and priority can be found using:
# tc filter show dev lo ingress
Another example replacing a filter (extending prior example):
/* We can also choose both (or one), let's try replacing an
* existing filter.
*/
DECLARE_LIBBPF_OPTS(bpf_tc_opts, replace_opts, .handle =
opts.handle, .priority = opts.priority,
.prog_fd = fd);
r = bpf_tc_attach(&hook, &replace_opts);
if (r == -EEXIST) {
/* Expected, now use BPF_TC_F_REPLACE to replace it */
replace_opts.flags = BPF_TC_F_REPLACE;
return bpf_tc_attach(&hook, &replace_opts);
} else if (r < 0) {
return r;
}
/* There must be no existing filter with these
* attributes, so cleanup and return an error.
*/
replace_opts.prog_fd = replace_opts.prog_id = 0;
bpf_tc_detach(&hook, &replace_opts);
return -1;
To obtain info of a particular filter:
/* Find info for filter with handle 1 and priority 50 */
DECLARE_LIBBPF_OPTS(bpf_tc_opts, info_opts, .handle = 1,
.priority = 50);
r = bpf_tc_query(&hook, &info_opts);
if (r == -ENOENT)
printf("Filter not found");
else if (r < 0)
return r;
printf("Prog ID: %u", info_opts.prog_id);
return 0;
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Co-developed-by: Daniel Borkmann <daniel@iogearbox.net> # libbpf API design
[ Daniel: also did major patch cleanup ]
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20210512103451.989420-3-memxor@gmail.com
This change introduces a few helpers to wrap open coded attribute
preparation in netlink.c. It also adds a libbpf_netlink_send_recv() that
is useful to wrap send + recv handling in a generic way. Subsequent patch
will also use this function for sending and receiving a netlink response.
The libbpf_nl_get_link() helper has been removed instead, moving socket
creation into the newly named libbpf_netlink_send_recv().
Every nested attribute's closure must happen using the helper
nlattr_end_nested(), which sets its length properly. NLA_F_NESTED is
enforced using nlattr_begin_nested() helper. Other simple attributes
can be added directly.
The maxsz parameter corresponds to the size of the request structure
which is being filled in, so for instance with req being:
struct {
struct nlmsghdr nh;
struct tcmsg t;
char buf[4096];
} req;
Then, maxsz should be sizeof(req).
This change also converts the open coded attribute preparation with these
helpers. Note that the only failure the internal call to nlattr_add()
could result in the nested helper would be -EMSGSIZE, hence that is what
we return to our caller.
The libbpf_netlink_send_recv() call takes care of opening the socket,
sending the netlink message, receiving the response, potentially invoking
callbacks, and return errors if any, and then finally close the socket.
This allows users to avoid identical socket setup code in different places.
The only user of libbpf_nl_get_link() has been converted to make use of it.
__bpf_set_link_xdp_fd_replace() has also been refactored to use it.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
[ Daniel: major patch cleanup ]
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20210512103451.989420-2-memxor@gmail.com
Static maps never really worked with libbpf, because all such maps were always
silently resolved to the very first map. Detect static maps (both legacy and
BTF-defined) and report user-friendly error.
Tested locally by switching few maps (legacy and BTF-defined) in selftests to
static ones and verifying that now libbpf rejects them loudly.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210513233643.194711-2-andrii@kernel.org
Daniel Borkmann says:
====================
pull-request: bpf 2021-05-11
The following pull-request contains BPF updates for your *net* tree.
We've added 13 non-merge commits during the last 8 day(s) which contain
a total of 21 files changed, 817 insertions(+), 382 deletions(-).
The main changes are:
1) Fix multiple ringbuf bugs in particular to prevent writable mmap of
read-only pages, from Andrii Nakryiko & Thadeu Lima de Souza Cascardo.
2) Fix verifier alu32 known-const subregister bound tracking for bitwise
operations and/or/xor, from Daniel Borkmann.
3) Reject trampoline attachment for functions with variable arguments,
and also add a deny list of other forbidden functions, from Jiri Olsa.
4) Fix nested bpf_bprintf_prepare() calls used by various helpers by
switching to per-CPU buffers, from Florent Revest.
5) Fix kernel compilation with BTF debug info on ppc64 due to pahole
missing TCP-CC functions like cubictcp_init, from Martin KaFai Lau.
6) Add a kconfig entry to provide an option to disallow unprivileged
BPF by default, from Daniel Borkmann.
7) Fix libbpf compilation for older libelf when GELF_ST_VISIBILITY()
macro is not available, from Arnaldo Carvalho de Melo.
8) Migrate test_tc_redirect to test_progs framework as prep work
for upcoming skb_change_head() fix & selftest, from Jussi Maki.
9) Fix a libbpf segfault in add_dummy_ksym_var() if BTF is not
present, from Ian Rogers.
10) Fix tx_only micro-benchmark in xdpsock BPF sample with proper frame
size, from Magnus Karlsson.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
For better future extensibility add per-file linker options. Currently
the set of available options is empty. This changes bpf_linker__add_file()
API, but it's not a breaking change as bpf_linker APIs hasn't been released
yet.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210507054119.270888-3-andrii@kernel.org
Pull networking fixes from Jakub Kicinski:
"Networking fixes for 5.13-rc1, including fixes from bpf, can and
netfilter trees. Self-contained fixes, nothing risky.
Current release - new code bugs:
- dsa: ksz: fix a few bugs found by static-checker in the new driver
- stmmac: fix frame preemption handshake not triggering after
interface restart
Previous releases - regressions:
- make nla_strcmp handle more then one trailing null character
- fix stack OOB reads while fragmenting IPv4 packets in openvswitch
and net/sched
- sctp: do asoc update earlier in sctp_sf_do_dupcook_a
- sctp: delay auto_asconf init until binding the first addr
- stmmac: clear receive all(RA) bit when promiscuous mode is off
- can: mcp251x: fix resume from sleep before interface was brought up
Previous releases - always broken:
- bpf: fix leakage of uninitialized bpf stack under speculation
- bpf: fix masking negation logic upon negative dst register
- netfilter: don't assume that skb_header_pointer() will never fail
- only allow init netns to set default tcp cong to a restricted algo
- xsk: fix xp_aligned_validate_desc() when len == chunk_size to avoid
false positive errors
- ethtool: fix missing NLM_F_MULTI flag when dumping
- can: m_can: m_can_tx_work_queue(): fix tx_skb race condition
- sctp: fix a SCTP_MIB_CURRESTAB leak in sctp_sf_do_dupcook_b
- bridge: fix NULL-deref caused by a races between assigning
rx_handler_data and setting the IFF_BRIDGE_PORT bit
Latecomer:
- seg6: add counters support for SRv6 Behaviors"
* tag 'net-5.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (73 commits)
atm: firestream: Use fallthrough pseudo-keyword
net: stmmac: Do not enable RX FIFO overflow interrupts
mptcp: fix splat when closing unaccepted socket
i40e: Remove LLDP frame filters
i40e: Fix PHY type identifiers for 2.5G and 5G adapters
i40e: fix the restart auto-negotiation after FEC modified
i40e: Fix use-after-free in i40e_client_subtask()
i40e: fix broken XDP support
netfilter: nftables: avoid potential overflows on 32bit arches
netfilter: nftables: avoid overflows in nft_hash_buckets()
tcp: Specify cmsgbuf is user pointer for receive zerocopy.
mlxsw: spectrum_mr: Update egress RIF list before route's action
net: ipa: fix inter-EE IRQ register definitions
can: m_can: m_can_tx_work_queue(): fix tx_skb race condition
can: mcp251x: fix resume from sleep before interface was brought up
can: mcp251xfd: mcp251xfd_probe(): add missing can_rx_offload_del() in error path
can: mcp251xfd: mcp251xfd_probe(): fix an error pointer dereference in probe
netfilter: nftables: Fix a memleak from userdata error path in new objects
netfilter: remove BUG_ON() after skb_header_pointer()
netfilter: nfnetlink_osf: Fix a missing skb_header_pointer() NULL check
...
One of our benchmarks running in (Google-internal) CI pushes data
through the ringbuf faster htan than userspace is able to consume
it. In this case it seems we're actually able to get >INT_MAX entries
in a single ring_buffer__consume() call. ASAN detected that cnt
overflows in this case.
Fix by using 64-bit counter internally and then capping the result to
INT_MAX before converting to the int return type. Do the same for
the ring_buffer__poll().
Fixes: bf99c936f9 (libbpf: Add BPF ring buffer support)
Signed-off-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210429130510.1621665-1-jackmanb@google.com
